# Does speeding up a diesel generator result in a better load response?

If I speed up my prime mover (500 kW CAT diesel generator) from 60 to 61 and 62 Hz 1800/1830/1860 rpm and hit it with 25/50/75/100 percent load steps, how much, if any kW/HP is gained?

I am simply trying to get more HP for a temporary transient about to hit the bus. How much will speeding up the gen set gain me? Anything?

• It depends on the equipment drawing the load, as well as the motor speed. Some equipment is frequency-sensitive, and might not take well to that change. Inductive loads, such as motors, might draw less current at higher frequency, which might be an additional benefit, or might prevent the motor from starting. Test, if you can. Jan 21 at 21:50
• Any halfway modern 500 kW generator is likely equipped with a whole bunch of protections that will trip if the load goes way out of spec. Those will likely be your main concern, unless you somehow managed to hack them all out. Jan 22 at 8:32
• Can you explain some more of the scenario with the type of load, and the transient for how long? Jan 22 at 11:47
• @Jeroen3 Randy explained it. The question is about load steps. A step means the power after the sudden increase/decrease stays constant. The question isn't about transient events exceeding nominal power. Jan 24 at 15:37

Higher RPM will proportionally raise the peak power a diesel engine can deliver - up to a point, of course.

But 62Hz is just 3% more than 60Hz, so it's not very relevant. You can get similar variation with intake air temperature, humidity and pressure changes.

What you really need is energy storage. On the mechanical side, that would be a direct-coupled or geared flywheel. On the mechano-electrical side, that would be an external motor-generator attached as a load, driving the flywheel. The motor/gen controller would dump the flywheel energy on line sags, and absorb it to re-spin the flywheel during light load.

• +1, but if we're really talking about a transient (say, split-second up to a couple of seconds), the extra rotational energy in the system could have a much bigger effect than 3%, because the generator likely doesn't wait until it stalls completely before dropping off the bus. Let's just make up some numbers and say it trips when the frequency drops below 55 Hz, then starting at 62 Hz would give you 40% more headroom for transients. Jan 22 at 8:27
• I think the question is not about peak power or sudden power excursions above nominal power. Randy asked about load steps happening within the normal power range. The governor of a genset isn't fast enough either to react on loadsteps and flatten out voltage drops. Jan 24 at 15:35

A genset has a significant rotating mass. The energy stored in rotating masses goes up with the square of the RPM.

In fact, if the genset runs at 3% higher speed it stocks 6.1 % extra mechanical energy. On a given load step the frequency drop will be smaller than that when running at nominal speed. However very short transient load changes are more likely to be served by the energy stored in the magnetic field. Without doing extensive calculations I think the inductivity will show a constant behaviour for these transient events i.e. independent from the genset speed.

However increasing frequency might have additional detrimental effects you should calculate carefully before considering a speed up as a measure.

Reactive power will most likely increase all over your network. At least all parasitic capacitors will show an increased current. As reactive power is eaten up partially by the ohmic resistance of the generator windings the benefit of the speed up might be eaten up partially by this effect.

I think eddy currents inside the generator might increase slightly as well. I have no idea how big that effect will be, most likely small, but there's some for sure.

• Thank you all very much for your input! Extremely helpful and insightful. My new thought is to raise VAC as well to the upper limits of acceptable level on my bus. I am actually going to run this test on a 1000 kw machine as well stand alone optimally tuned for critical dampening then combine machines n parallel and repeat tests. But man thank you all. I at least have a ball park number 6% ish to target. V/r Jan 25 at 15:15
• The end goal / transient is LASER BEAMS. Buffered with a cap bank And batteries, BIG ONES! Jan 25 at 15:44